Progress in FY2018 was in the following areas: (1) RAPID MIXING AND FREEZING TECHNOLOGY FOR SSNMR. We have perfected an apparatus for rapid mixing and freezing that will enable studies of transient intermediates on the millisecond time scale. We have improved our mixer design so that full mixing of 20% glycerol solutions occurs within 2 ms. The time before freezing can then be varied from 0.5 ms to 20 ms or longer, by varying the path length between the mixer and the rotating cold copper surface used to freeze solutions that leave the mixer as a high-speed jet. Various experiments are planned, including studies of the initial stages of self-assembly of the HIV-1 capsid protein to form an ordered protein lattice, which will involve rapid mixing of an unassembled capsid protein solution with high-salt buffer to initiate the lattice assembly process. In experiments so far, we have examined the folding and tetramerization process of the bee venom peptide melittin, following a rapid pH jump (from pH 3 to pH 7) that initiates the process in our mixer. At the shortest time (0.5 ms delay between mixing and freezing), 2D solid state NMR spectra show that melittin is primarily non-helical. At longest time (20 ms delay), melittin is primarily helical, with intermolecular contacts consistent with the known tetramer structure of the fully folded state. 2D spectra at intermediate times allow us to map out the time dependence of the folding/tetramerization process. A manuscript describing this work will be submitted later in 2018. (2) MICRON-SCALE MRI: We have obtained the first high-resolution 3D magnetic resonance images of test samples at low temperatures (15-30 K) range. The signal-to-noise enhancement that arises from enhanced nuclear spin polarizations at low temperatures allows us to achieve isotropic resolution of 2.8 microns over a total volume of 1.5 nanoliters, using a sample of 20 micron-diameter glass beads in paramagnetically doped glycerol/water. This represents an approximately 30% improvement over the best MRI resolution previously reported previously in the literature. This work has now been published (Chen and Tycko, J. Magn. Reson. 2018). We are now working to improve the resolution by using dynamic nuclear polarization (DNP) to enhance NMR signal strengths. We have found that the nitroxide-based dopants that are required for DNP produce significant reductions in NMR relaxation times (T1rho relaxation during the pulsed spin locking signal detection period of our MRI pulse sequence and T2 relaxation during Lee-Goldburg homonuclear decoupling during the phase-encoded imaging period). At temperatures above 15 K, these reductions in relaxation times result in signal reductions that cancel the signal enhancements from DNP. However, we have found that the relaxation times become sufficiently long again when the sample temperature is reduced to <10 K, indicating that DNP-enhanced MRI should be successful at sufficiently low temperatures. A manuscript describing the temperature dependence of NMR relaxation in nitroxide-doped samples, in the 3-30 K range, has been submitted for publication. This paper includes a theoretical explanation, based on the principle that NMR relaxation is caused by local magnetic field fluctuations from electron spin flip-flop transitions, and that these flip-flop transitions are suppressed when the sample temperature become so low that electron spins are highly polarized at thermal equilibrium (i.e., thermal energy is less than the electron spin-flip energy in the NMR magnet).